CN110310474A - A kind of vehicle flowrate prediction technique and device based on space-time residual error network - Google Patents
A kind of vehicle flowrate prediction technique and device based on space-time residual error network Download PDFInfo
- Publication number
- CN110310474A CN110310474A CN201810457554.2A CN201810457554A CN110310474A CN 110310474 A CN110310474 A CN 110310474A CN 201810457554 A CN201810457554 A CN 201810457554A CN 110310474 A CN110310474 A CN 110310474A
- Authority
- CN
- China
- Prior art keywords
- data
- space
- residual error
- error network
- wagon flow
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/065—Traffic control systems for road vehicles by counting the vehicles in a section of the road or in a parking area, i.e. comparing incoming count with outgoing count
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/123—Traffic control systems for road vehicles indicating the position of vehicles, e.g. scheduled vehicles; Managing passenger vehicles circulating according to a fixed timetable, e.g. buses, trains, trams
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
Abstract
The invention discloses a kind of vehicle flowrate prediction technique and device based on space-time residual error network, this method obtains the geographic latitude and longitude in city to be measured, the city is mapped as to the grid of an I*J according to the geographic latitude and longitude of acquisition according to preset ratio, wherein each grid represents a region；The vehicle driving trace of each period is counted by what is obtained using GPS, and respective record, the data collected are carried out to the weather conditions of acquisition；Pretreatment is normalized to the collected data；Space-time residual error network is trained according to pretreated data, wagon flow prediction model has been trained in formation；The prediction model is called to predict the vehicle flowrate and assessment prediction error of a region particular moment.Using the present invention carry out vehicle flowrate prediction, by be based on space-time residual error network wagon flow prediction model, influenced each other using interregional, can to data carry out deeper into mining analysis, largely also further improve wagon flow prediction accuracy.
Description
Technical field
The present invention relates to the technical fields of deep learning and intelligent vehicle system transposition, in particular to a kind of to be based on space-time residual error
The vehicle flowrate prediction technique and device of network.
Background technique
With the development and improvement of living standard of society, city vehicle increasingly congestion, vehicle accident and air pollution also into
The aggravation of one step.For more accurate pre- measuring car congestion, more reasonable vehicle route planning is provided for vehicle driving, is had very much
Necessity carries out large-scale vehicle network congestion prediction.
In vehicle network, the vehicle congestion of road is interactional, the vehicle congestion status in a region and adjacent
The congestion status in region is inseparable, so predicting that each region vehicle dynamic changes need to consider from global network；Moreover, right
Single section, which carries out vehicle congestion prediction, has short-sighted property, and local vehicle prediction only only by historical data or is based on the limited section in periphery
The state of vehicle is predicted, but when section is expanded to extensive vehicle neural network forecast, operand can be greatly increased, forecasting efficiency
Guarantee is all unable to get with precision, it is impossible to meet the real-time of vehicle information service and accuracys.
Applicant of the present invention by the prior art document carry out largely retrieve after find, existing vehicle prediction side
Method has that low efficiency and precision are low.Therefore, how to solve existing method cannot trans-regional and permanent memory history wagon flow
Data become so that this problem of vehicle flowrate can not be more accurately predicted according to the input between adjacent area, output vehicle flowrate
For the important research direction of those skilled in the art.
Summary of the invention
The present invention provides a kind of vehicle flowrate prediction technique and dress based on space-time residual error network according to above-mentioned technical background
It sets, it is therefore an objective to by being based on space-time residual error network wagon flow prediction model, be influenced each other using interregional, obtain preferably prediction
As a result.
In a first aspect, the present invention provides a kind of vehicle flowrate prediction technique based on space-time residual error network, including following step
It is rapid:
The geographic latitude and longitude for obtaining city to be measured, by the city according to the geographic latitude and longitude of acquisition according to preset ratio
It is mapped as the grid of an I*J, wherein each grid represents a region；
The vehicle driving trace of each period is counted by what is obtained using GPS, and the weather conditions to acquisition
Carry out respective record, the data collected；
Pretreatment is normalized to the collected data；
Space-time residual error network is trained according to pretreated data, wagon flow prediction model has been trained in formation；
The prediction model is called to predict the vehicle flowrate and assessment prediction error of a region particular moment.
Preferably, the vehicle driving trace of each period is counted by what is obtained using GPS, and to the day of acquisition
Vaporous condition carries out respective record, the specific steps for the data collected are as follows: will utilize vehicle in each period of GPS statistics
It is recorded in a region and its input of adjacent area, output trajectory, and records the special event in each period, day
Gas；Several related datas in each period are inputted respectively, is exported and is summarized and polymerize.
Preferably, pretreated specific steps are normalized to the collected data are as follows: by the input, defeated after polymerization
Data are normalized out, realize especially by following formula:
Wherein x*To normalize pretreated data, xmFor sample data minimum value, xMFor sample data maximum value, x is
Wait normalize pretreated history wagon flow data；The sample data specifically refers to: for training and all data tested.
Preferably, space-time residual error network is trained according to pretreated data, wagon flow prediction model has been trained in formation
Specific steps are as follows: the parameters of clock synchronization sky residual error Network Prediction Model carry out reasonable set, and the parameter includes: input layer
Number, middle layer, weight factor, layer output and input vector dimension, the nodal point number and output node number of each hidden layer；According to pre-
Processed data are trained the space-time residual error network model that Reasonable Parameters have been arranged, and formation has trained wagon flow to predict mould
Type.
Preferably, according to pretreated data, the space-time residual error network model that Reasonable Parameters have been arranged is trained,
Form the specific steps for having trained wagon flow prediction model are as follows: training will be divided by polymerizeing and normalizing pretreated data
Data set and test data set；According to training dataset, in the space-time residual error network wagon flow prediction model for having set Reasonable Parameters
It is upper to be trained using back-propagation algorithm, obtain the normalization prediction of the corresponding wagon flow data to next specified time interval
Value；The normalization predicted value of the wagon flow data at next specified time interval is subjected to anti-normalization processing, is obtained to specified
The wagon flow data predicted value of time interval.
Preferably, the prediction model is called to predict the vehicle flowrate of a region particular moment and the tool of assessment prediction error
Body step are as follows: according to the wagon flow for having trained wagon flow prediction model prediction specified time interval, and assessment prediction error.
Second aspect, the present invention provides a kind of vehicle flowrate prediction meanss based on space-time residual error network, comprising:
Area maps conversion module: for obtaining the geographic latitude and longitude in a city, by the city according to the ground of acquisition
Reason longitude and latitude is mapped as the grid of an I*J according to a certain percentage, and wherein each grid represents a region；
Data acquisition module: for being counted what is obtained using GPS to the vehicle driving trace of each period, and
Respective record, the data collected are carried out to the weather conditions of acquisition；
Data processing module: for pretreatment to be normalized to the collected data；
Training pattern module: for according to pretreated data, to the space-time residual error network mould that Reasonable Parameters have been arranged
Type is trained, and wagon flow prediction model has been trained in formation；
Assessment prediction error module: for calling the prediction model to predict the vehicle flowrate of a region particular moment and commenting
Estimate prediction error.
Preferably, the above-mentioned vehicle flowrate prediction meanss based on space-time residual error network, further includes: space-time residual error neural network forecast mould
Shape parameter setting module: reasonable set is carried out for the parameters to space-time residual error Network Prediction Model.
Preferably, training pattern module, comprising:
Division module: training dataset and test number are divided into for that will pass through polymerization and normalize pretreated data
According to collection；
Training module: for predicting mould in the space-time residual error network wagon flow for having set Reasonable Parameters according to training dataset
It is trained in type using back-propagation algorithm, the normalization for obtaining the corresponding wagon flow data to next specified time interval is pre-
Measured value；
Processing module: for the normalization predicted value of the wagon flow data at next specified time interval to be carried out anti-normalizing
Change processing, obtains the wagon flow data predicted value to specified time interval.
The beneficial effect of above-mentioned technical proposal provided by the invention includes at least:
Compared with prior art, the present invention provides a kind of vehicle flowrate prediction technique and dress based on space-time residual error network
It sets, study is trained to wagon flow data by space-time residual error network model, vehicle flowrate data can be carried out more profound
Study；Compared with traditional wagon flow data prediction model, data can be carried out with more deep analysis, and effectively extract number
According to potential layered characteristic, improve the efficiency of feature extraction.Therefore vehicle flowrate prediction is carried out using prediction model of the invention, it can
With to data carry out deeper into mining analysis, largely also further improve wagon flow prediction accuracy.
Other features and advantages of the present invention will be illustrated in the following description, also, partly becomes from specification
It obtains it is clear that understand through the implementation of the invention.The objectives and other advantages of the invention can be by written explanation
Specifically noted structure is achieved and obtained in book, claims and attached drawing.
Below by the drawings and specific embodiments, technical scheme of the present invention will be described in further detail.
Detailed description of the invention
Attached drawing is used to provide further understanding of the present invention, and constitutes part of specification, with reality of the invention
It applies example to be used to explain the present invention together, not be construed as limiting the invention.In the accompanying drawings:
Fig. 1 is the flow chart of the vehicle amount prediction technique provided in an embodiment of the present invention based on space-time residual error network；
Fig. 2 is the wagon flow spirogram provided in an embodiment of the present invention for calculating city each region to be measured and inputting, exporting；
Fig. 3 is the structural block diagram of space-time residual error Network Prediction Model provided in an embodiment of the present invention；
Fig. 4 is the block diagram of the vehicle amount prediction meanss provided in an embodiment of the present invention based on space-time residual error network.
Specific embodiment
Exemplary embodiments of the present disclosure are described in more detail below with reference to accompanying drawings.Although showing the disclosure in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
It is fully disclosed to those skilled in the art.
Embodiment:
Referring to Fig.1 shown in -2, the vehicle flowrate prediction technique provided in an embodiment of the present invention based on space-time residual error network, including
Following steps:
1) area maps convert: the geographic latitude and longitude in city to be measured are obtained, by city to be measured according to the geographical longitude and latitude of acquisition
Degree is mapped as the grid of an I*J according to preset ratio, and wherein each grid represents a region.
2) data acquisition and processing (DAP): counting the vehicle driving trace of each period for what is obtained using GPS, and
Respective record, the data collected are carried out to the weather conditions of acquisition；Pretreatment is normalized to the data of collection.
S21, the acquisition to wagon flow data: by vehicle in each period counted using GPS in a region and its phase
The input in neighbouring region, output trajectory are recorded, and record the special event in each period, weather；By each period
Several interior related datas are inputted respectively, export and summarize and polymerize.
Specifically:
It is located at the grid (i, j) of i row j column for a region, in the input of time t momentOutputWagon flow
Amount data set can respectively indicate are as follows:
Wherein P is indicated in time interval tthThe vehicle driving trace of collection, Tr:g1→g2→........→g|Tr|It is in P
A motion profile, gkIt is geospatial coordinates；gk∈ (i, j) indicates point gkIn grid (i, j).
In time interval tth, in I × J region, all inputs, the vehicle flowrate exported can be expressed as a tensor
Xt∈R2×I×J, whereinFor example, if a region at a certain time interval
The vehicle of input is 3, and output is 1 and can be expressed as f (3,1).
S22, the pretreatment to wagon flow data: pretreatment is normalized to the data after polymerization.
Pretreatment specifically is normalized to data to realize by following formula:
Wherein x*To normalize pretreated data, xmFor sample data minimum value, xMFor sample data maximum value, x is
Wait normalize pretreated history wagon flow data.
3) training pattern and vehicle flowrate is predicted: space-time residual error network is trained according to pretreated data,
Wagon flow prediction model has been trained in formation；Prediction model is called to predict the vehicle flowrate and assessment prediction mistake of a region particular moment
Difference.
The parameters of S31, reasonable set space-time residual error Network Prediction Model.
Specifically:
If the historical perspective data set of input are as follows: X={ X0,X1,.......,Xn-1}；The data characteristics that external factor is extracted
Are as follows: E={ E0,E1........En-1}；The time interval of three time slices is respectively adjacent moment lc, time interval farther out
Same time point lp, farther time interval lq.The period q of the period p of time interval, farther time interval farther out.
Referring to the prediction model of space-time residual error network shown in Fig. 3, data are temporally divided into three timeslices by we first
Data in three time slices are first input into convolutional neural networks as a research attribute and are followed by residual error list by section
In the identical network model framework of metasequence, this model structure can make adjacent area and have space independent compared with far region
Property.Different attributes is arranged to the output in different attributes and region；
This prediction model is made of conv1, conv2 and L residual unit ResUnit of two convolutional layers, conv1 and all
Residual unit be 64 convolution kernel sizes be 3*3 Feature Mapping layer, the spy that 2 convolution kernel sizes that conv2 is are 3*3
Mapping layer is levied, batch size is 32.
S32, according to pretreated data, the space-time residual error network model that Reasonable Parameters have been arranged is trained, shape
At having trained wagon flow prediction model.
S32.1, training dataset and test data set will be divided by polymerizeing and normalizing pretreated data.
S32.2, according to training dataset, carried out on the space-time residual error network wagon flow prediction model for having set Reasonable Parameters
Training, and space-time residual error Network Prediction Model is calculated in the error of test data set, detailed process are as follows:
The data set of input is divided into three time slices by the historical perspective data set X of input, is separately input to respectively
Network structure be trained, the result after training is polymerize；External factor is input to the net there are two Quan Lian stratum
Network structure, carries out feature extraction, and the feature of extraction is E；The result of external factor and the result of three time attributes are carried out again
Prediction wagon flow data result is calculated in polymerization.
Assuming that the fragmentation of data of adjacent moment segment isIt will count
It is attached to form a tensor using time interval as axis according to fragmentFirst willPass through a convolution mind
It is obtained after network:
Wherein * indicates convolution algorithm, and f is an excitation function, such as rectifier f (z)=max (0, z)；Wc (1),It is
The learning parameter of first layer.
Define the Feature Mapping of residual unit are as follows:
X(l+1)=X(l)+F(X(l)), l=1,2,3 ... ..., L (2)
Wherein X(l)And X(l+1)It is outputting and inputting for l layers of residual unit respectively.
By the data by conv1 output using L layers of residual unit, every layer of residual unit are as follows:
Wherein F is a residual error function, θ(l)All learning parameters for including for first of residual unit.Finally again through one
Convolutional neural networks conv2 obtains the output of adjacent moment
Using identical method, we can calculate lp、lqSame time point output, respectively to wagon flow data
Periodically predicted with tendency.Assuming that the time interval that the period is p is lp, so the intimate correlated series in period areIt is a residual by the convolution operation of formula (1) and the L of formula (3)
Poor unit operation, the period output for obtaining the moment farther out areAlso the trend at available farther moment, which exports, is
The input data set of trend(1≤t≤n-1), wherein lqIt is between trend data collection
Every q is the period of trend data collection.P and q is the period of two different types, and p is one day in the specific implementation, and q is one
Week.
The output of three attributes is polymerize, different attributes has different weights, the output after polymerization are as follows:
Wherein ο is Hadamard convolution, Wc、Wp、WqBe respectively adjust neighbouring time interval, farther out time interval, it is farther when
Between be spaced three attributes weight learning parameter.
Influence for the factor of external factor extracts some features from external factor data manually, such as weather conditions,
The factors such as special event.External factor data are obtained into output X by two fully connected network network layersExt, external factor is influenced
Output and the output of three time attributes are polymerize, and obtain the pre- measuring car in t-th of time interval using Tanh activation primitive
Flow value:
Wherein tanh is a hyperbolic tangent function, it can be ensured that the value of output is between [- 1,1].
By the space-time residual error network model trained, the history that can be influenced using three time attributes and outer not factor
Wagon flow data are to XtThe prediction for carrying out wagon flow data, obtains predicted valueObjective function in training process is predicted value and true
Real value obtains root-mean-square error:
Wherein θ is all learning parameters of residual error network, and Xt is actual vehicle flowrate,It is the vehicle flowrate of prediction.
S32.3, the normalization predicted value of the wagon flow data at the obtained next specified time interval step S32.2 is carried out instead
Normalized obtains the wagon flow data predicted value to specified time interval.
S33, basis have trained the wagon flow at wagon flow prediction model prediction specified time interval, and assessment prediction error；It is square
Root error (RMSE) are as follows:
WhereinIt is the value of prediction, x is the practical vehicle flowrate in region.
Based on the same inventive concept, the embodiment of the invention also provides the vehicle flowrates based on space-time residual error network to predict dress
It sets, since the principle of the solved problem of the device is similar to the aforementioned vehicle flowrate prediction technique based on space-time residual error network,
The implementation of the device may refer to the implementation of preceding method, and overlaps will not be repeated.
The embodiment of the invention also provides a kind of vehicle flowrate prediction meanss based on space-time residual error network, referring to shown in Fig. 4,
Include:
Area maps conversion module 41: for obtaining the geographic latitude and longitude in a city, by the city according to the ground of acquisition
Reason longitude and latitude is mapped as the grid of an I*J according to a certain percentage, and wherein each grid represents a region；
Data acquisition module 42: for being counted what is obtained using GPS to the vehicle driving trace of each period,
And respective record, the data collected are carried out to the weather conditions of acquisition；
Data processing module 43: for pretreatment to be normalized to the collected data；
Training pattern module 45: for according to pretreated data, to the space-time residual error network that Reasonable Parameters have been arranged
Model is trained, and wagon flow prediction model has been trained in formation；
Assessment prediction error module 46: for calling the prediction model to predict the vehicle flowrate of a region particular moment simultaneously
Assessment prediction error.
In one embodiment, the above-mentioned vehicle flowrate prediction meanss based on space-time residual error network are also wrapped referring to shown in Fig. 4
It includes:
Space-time residual error Network Prediction Model parameter setting module 44: for each ginseng to space-time residual error Network Prediction Model
Number carries out reasonable set.
In one embodiment, above-mentioned training pattern module 45, referring to shown in Fig. 4, comprising:
Division module 451: training dataset and survey are divided into for that will pass through polymerization and normalize pretreated data
Try data set；
Training module 452: for being predicted in the space-time residual error network wagon flow for having set Reasonable Parameters according to training dataset
It is trained on model using back-propagation algorithm, obtains the normalization of the corresponding wagon flow data to next specified time interval
Predicted value；
Processing module 453: for the normalization predicted value of the wagon flow data at next specified time interval to be carried out anti-normalizing
Change processing, obtains the wagon flow data predicted value to specified time interval.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The shape for the computer program product implemented in usable storage medium (including but not limited to magnetic disk storage and optical memory etc.)
Formula.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
Mind and range.In this way, if these modifications and changes of the present invention belongs to the range of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to include these modifications and variations.
Claims (9)
1. a kind of vehicle flowrate prediction technique based on space-time residual error network, it is characterised in that: the following steps are included:
The geographic latitude and longitude for obtaining city to be measured maps in the city according to the geographic latitude and longitude of acquisition according to preset ratio
For the grid of an I*J, wherein each grid represents a region；
The vehicle driving trace of each period is counted by what is obtained using GPS, and the weather conditions of acquisition are carried out
Respective record, the data collected；
Pretreatment is normalized to the collected data；
Space-time residual error network is trained according to pretreated data, wagon flow prediction model has been trained in formation；
The prediction model is called to predict the vehicle flowrate and assessment prediction error of a region particular moment.
2. the vehicle flowrate prediction technique according to claim 1 based on space-time residual error network, it is characterised in that: will utilize
What GPS was obtained counts the vehicle driving trace of each period, and carries out respective record to the weather conditions of acquisition, obtains
To the specific steps of the data of collection are as follows:
Vehicle in each period counted using GPS is carried out in the input of a region and its adjacent area, output trajectory
Record, and record the special event in each period, weather；By several related datas difference in each period
It inputted, export and summarize and polymerize.
3. the vehicle flowrate prediction technique according to claim 2 based on space-time residual error network, it is characterised in that: to the receipts
Pretreated specific steps are normalized in the data of collection are as follows:
Input after polymerization, output data are normalized, realized especially by following formula:
Wherein x*To normalize pretreated data, xmFor sample data minimum value, xMFor sample data maximum value, x is wait return
One changes pretreated history wagon flow data；The sample data specifically refers to: for training and all data tested.
4. the vehicle flowrate prediction technique according to claim 1 based on space-time residual error network, it is characterised in that: according to pre- place
The data of reason are trained space-time residual error network, form the specific steps for having trained wagon flow prediction model are as follows:
Reasonable set is carried out to the parameters of space-time residual error Network Prediction Model, the parameter includes: the input number of plies, intermediate
Layer, weight factor, layer output and input vector dimension, the nodal point number and output node number of each hidden layer；
According to pretreated data, the space-time residual error network model that Reasonable Parameters have been arranged is trained, formation has been trained
Wagon flow prediction model.
5. the vehicle flowrate prediction technique according to claim 4 based on space-time residual error network, it is characterised in that: according to pre- place
The data managed are trained the space-time residual error network model that Reasonable Parameters have been arranged, and wagon flow prediction model has been trained in formation
Specific steps are as follows:
Training dataset and test data set will be divided by polymerizeing and normalizing pretreated data；
According to training dataset, calculated on the space-time residual error network wagon flow prediction model for set Reasonable Parameters using backpropagation
Method is trained, and obtains the normalization predicted value of the corresponding wagon flow data to next specified time interval；
The normalization predicted value of the wagon flow data at next specified time interval is subjected to anti-normalization processing, is obtained to specified
The wagon flow data predicted value of time interval.
6. the vehicle flowrate prediction technique according to claim 4 based on space-time residual error network, it is characterised in that: described in calling
Prediction model predicts the vehicle flowrate of a region particular moment and the specific steps of assessment prediction error are as follows: according to having trained wagon flow
Prediction model predicts the wagon flow at specified time interval, and assessment prediction error.
7. a kind of vehicle flowrate prediction meanss based on space-time residual error network characterized by comprising
Area maps conversion module: for obtaining the geographic latitude and longitude in a city, the city is passed through according to the geography of acquisition
Latitude is mapped as the grid of an I*J according to a certain percentage, and wherein each grid represents a region；
Data acquisition module: for being counted what is obtained using GPS to the vehicle driving trace of each period, and to obtaining
The weather conditions taken carry out respective record, the data collected；
Data processing module: for pretreatment to be normalized to the collected data；
Training pattern module: for according to pretreated data, to be arranged the space-time residual error network models of Reasonable Parameters into
Wagon flow prediction model has been trained in row training, formation；
Assessment prediction error module: for calling the prediction model to predict the vehicle flowrate of a region particular moment and assessing pre-
Survey error.
8. the vehicle flowrate prediction meanss according to claim 7 based on space-time residual error network, which is characterized in that further include:
Space-time residual error Network Prediction Model parameter setting module: it is reasonable to carry out for the parameters to space-time residual error Network Prediction Model
Setting.
9. the vehicle flowrate prediction meanss according to claim 7 or 8 based on space-time residual error network, which is characterized in that training
Model module includes:
Division module: training dataset and test data are divided into for that will pass through polymerization and normalize pretreated data
Collection；
Training module: it is used for according to training dataset, on the space-time residual error network wagon flow prediction model for having set Reasonable Parameters
It is trained using back-propagation algorithm, obtains the normalization prediction of the corresponding wagon flow data to next specified time interval
Value；
Processing module: for carrying out the normalization predicted value of the wagon flow data at next specified time interval at renormalization
Reason, obtains the wagon flow data predicted value to specified time interval.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810457554.2A CN110310474A (en) | 2018-05-14 | 2018-05-14 | A kind of vehicle flowrate prediction technique and device based on space-time residual error network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810457554.2A CN110310474A (en) | 2018-05-14 | 2018-05-14 | A kind of vehicle flowrate prediction technique and device based on space-time residual error network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110310474A true CN110310474A (en) | 2019-10-08 |
Family
ID=68073924
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810457554.2A Pending CN110310474A (en) | 2018-05-14 | 2018-05-14 | A kind of vehicle flowrate prediction technique and device based on space-time residual error network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110310474A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111130890A (en) * | 2019-12-26 | 2020-05-08 | 深圳市高德信通信股份有限公司 | Network flow dynamic prediction system |
| CN111197991A (en) * | 2020-01-15 | 2020-05-26 | 西安电子科技大学 | Method for predicting optimal driving path of vehicle based on deep neural network |
| CN111260121A (en) * | 2020-01-12 | 2020-06-09 | 桂林电子科技大学 | Urban-range pedestrian flow prediction method based on deep bottleneck residual error network |
| CN111540196A (en) * | 2020-03-25 | 2020-08-14 | 武汉纵横智慧城市股份有限公司 | Traffic flow early warning method, device, storage medium and device based on high-point video |
| CN111832876A (en) * | 2020-03-30 | 2020-10-27 | 北京骑胜科技有限公司 | Vehicle scheduling method, readable storage medium and electronic device |
| CN112418521A (en) * | 2020-11-23 | 2021-02-26 | 青岛科技大学 | Short-term marine fish school and fish quantity prediction method |
| CN112466117A (en) * | 2020-11-24 | 2021-03-09 | 南通大学 | Road network short-term traffic flow prediction method based on deep space-time residual error network |
| CN112766240A (en) * | 2021-03-18 | 2021-05-07 | 福州大学 | Residual multi-graph convolution crowd distribution prediction method and system based on space-time relationship |
| CN112907971A (en) * | 2021-02-04 | 2021-06-04 | 南通大学 | Urban road network short-term traffic flow prediction method based on genetic algorithm optimization space-time residual error model |
| CN113095535A (en) * | 2020-01-08 | 2021-07-09 | 普天信息技术有限公司 | Flow prediction method and device based on deep space-time residual error network |
| CN113205685A (en) * | 2021-04-30 | 2021-08-03 | 南通大学 | Short-term traffic flow prediction method based on global-local residual error combination model |
| CN113327417A (en) * | 2021-05-28 | 2021-08-31 | 南通大学 | Traffic flow prediction method based on 3D dynamic space-time residual convolution associated network |
| WO2021212866A1 (en) * | 2020-04-21 | 2021-10-28 | 长安大学 | Vehicle travel volume prediction model construction method, and prediction method and system |
| CN113610059A (en) * | 2021-09-13 | 2021-11-05 | 北京百度网讯科技有限公司 | Vehicle control method and device based on regional assessment and intelligent traffic management system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103021191A (en) * | 2012-11-22 | 2013-04-03 | 浙江理工大学 | Intelligent traffic control device and control method |
| CN103198672A (en) * | 2013-03-27 | 2013-07-10 | 大连海事大学 | Method for laying urban road network traffic flow detectors |
| CN105513354A (en) * | 2015-12-22 | 2016-04-20 | 电子科技大学 | Video-based urban road traffic jam detecting system |
| CN105654729A (en) * | 2016-03-28 | 2016-06-08 | 南京邮电大学 | Short-term traffic flow prediction method based on convolutional neural network |
| CN107103758A (en) * | 2017-06-08 | 2017-08-29 | 厦门大学 | A kind of city area-traffic method for predicting based on deep learning |
| CN107730887A (en) * | 2017-10-17 | 2018-02-23 | 海信集团有限公司 | Realize method and device, the readable storage medium storing program for executing of traffic flow forecasting |
| US20180096594A1 (en) * | 2011-11-22 | 2018-04-05 | Fastec International, Llc | Systems and methods involving features of adaptive and/or autonomous traffic control |
-
2018
- 2018-05-14 CN CN201810457554.2A patent/CN110310474A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180096594A1 (en) * | 2011-11-22 | 2018-04-05 | Fastec International, Llc | Systems and methods involving features of adaptive and/or autonomous traffic control |
| CN103021191A (en) * | 2012-11-22 | 2013-04-03 | 浙江理工大学 | Intelligent traffic control device and control method |
| CN103198672A (en) * | 2013-03-27 | 2013-07-10 | 大连海事大学 | Method for laying urban road network traffic flow detectors |
| CN105513354A (en) * | 2015-12-22 | 2016-04-20 | 电子科技大学 | Video-based urban road traffic jam detecting system |
| CN105654729A (en) * | 2016-03-28 | 2016-06-08 | 南京邮电大学 | Short-term traffic flow prediction method based on convolutional neural network |
| CN107103758A (en) * | 2017-06-08 | 2017-08-29 | 厦门大学 | A kind of city area-traffic method for predicting based on deep learning |
| CN107730887A (en) * | 2017-10-17 | 2018-02-23 | 海信集团有限公司 | Realize method and device, the readable storage medium storing program for executing of traffic flow forecasting |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111130890A (en) * | 2019-12-26 | 2020-05-08 | 深圳市高德信通信股份有限公司 | Network flow dynamic prediction system |
| CN113095535A (en) * | 2020-01-08 | 2021-07-09 | 普天信息技术有限公司 | Flow prediction method and device based on deep space-time residual error network |
| CN111260121A (en) * | 2020-01-12 | 2020-06-09 | 桂林电子科技大学 | Urban-range pedestrian flow prediction method based on deep bottleneck residual error network |
| CN111260121B (en) * | 2020-01-12 | 2022-04-29 | 桂林电子科技大学 | Urban-range pedestrian flow prediction method based on deep bottleneck residual error network |
| CN111197991A (en) * | 2020-01-15 | 2020-05-26 | 西安电子科技大学 | Method for predicting optimal driving path of vehicle based on deep neural network |
| CN111197991B (en) * | 2020-01-15 | 2022-09-23 | 西安电子科技大学 | Method for predicting optimal driving path of vehicle based on deep neural network |
| CN111540196A (en) * | 2020-03-25 | 2020-08-14 | 武汉纵横智慧城市股份有限公司 | Traffic flow early warning method, device, storage medium and device based on high-point video |
| CN111832876A (en) * | 2020-03-30 | 2020-10-27 | 北京骑胜科技有限公司 | Vehicle scheduling method, readable storage medium and electronic device |
| WO2021212866A1 (en) * | 2020-04-21 | 2021-10-28 | 长安大学 | Vehicle travel volume prediction model construction method, and prediction method and system |
| CN112418521B (en) * | 2020-11-23 | 2023-02-24 | 青岛科技大学 | Short-term marine fish school and fish quantity prediction method |
| CN112418521A (en) * | 2020-11-23 | 2021-02-26 | 青岛科技大学 | Short-term marine fish school and fish quantity prediction method |
| CN112466117A (en) * | 2020-11-24 | 2021-03-09 | 南通大学 | Road network short-term traffic flow prediction method based on deep space-time residual error network |
| CN112907971A (en) * | 2021-02-04 | 2021-06-04 | 南通大学 | Urban road network short-term traffic flow prediction method based on genetic algorithm optimization space-time residual error model |
| CN112907971B (en) * | 2021-02-04 | 2022-06-10 | 南通大学 | Urban road network short-term traffic flow prediction method based on genetic algorithm optimization space-time residual error model |
| CN112766240A (en) * | 2021-03-18 | 2021-05-07 | 福州大学 | Residual multi-graph convolution crowd distribution prediction method and system based on space-time relationship |
| CN112766240B (en) * | 2021-03-18 | 2022-08-12 | 福州大学 | Residual multi-graph convolution crowd distribution prediction method and system based on space-time relationship |
| CN113205685A (en) * | 2021-04-30 | 2021-08-03 | 南通大学 | Short-term traffic flow prediction method based on global-local residual error combination model |
| CN113327417B (en) * | 2021-05-28 | 2022-06-10 | 南通大学 | Traffic flow prediction method based on 3D dynamic space-time residual convolution associated network |
| CN113327417A (en) * | 2021-05-28 | 2021-08-31 | 南通大学 | Traffic flow prediction method based on 3D dynamic space-time residual convolution associated network |
| CN113610059A (en) * | 2021-09-13 | 2021-11-05 | 北京百度网讯科技有限公司 | Vehicle control method and device based on regional assessment and intelligent traffic management system |
| CN113610059B (en) * | 2021-09-13 | 2023-12-05 | 北京百度网讯科技有限公司 | Vehicle control method and device based on regional assessment and intelligent traffic management system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110310474A (en) | A kind of vehicle flowrate prediction technique and device based on space-time residual error network | |
| CN110264709B (en) | Method for predicting traffic flow of road based on graph convolution network | |
| CN110570651B (en) | Road network traffic situation prediction method and system based on deep learning | |
| CN106981198B (en) | Deep learning network model for travel time prediction and establishing method thereof | |
| CN104330089B (en) | A kind of method that map match is carried out using history gps data | |
| CN109840660A (en) | A kind of vehicular characteristics data processing method and vehicle risk prediction model training method | |
| CN104217002B (en) | A kind of road condition information filling method obtained based on quality data | |
| CN109871876A (en) | A kind of Freeway Conditions identification and prediction technique based on floating car data | |
| CN111696355A (en) | Dynamic graph convolution traffic speed prediction method | |
| CN105513370B (en) | The traffic zone division methods excavated based on sparse license plate identification data | |
| CN104764868B (en) | A kind of soil organic matter Forecasting Methodology based on Geographical Weighted Regression | |
| CN102610092A (en) | Urban road speed predication method based on RBF (radial basis function) neural network | |
| CN109615860A (en) | A kind of signalized intersections method for estimating state based on nonparametric Bayes frame | |
| CN113902580B (en) | Historical farmland distribution reconstruction method based on random forest model | |
| CN114944053A (en) | Traffic flow prediction method based on spatio-temporal hypergraph neural network | |
| CN104850677A (en) | Deep learning cellular automaton simulation analysis method based on geographic information system | |
| CN113327417B (en) | Traffic flow prediction method based on 3D dynamic space-time residual convolution associated network | |
| Mirzahossein et al. | How realistic is static traffic assignment? Analyzing automatic number-plate recognition data and image processing of real-time traffic maps for investigation | |
| CN105243503A (en) | Coastal zone ecological safety assessment method based on space variables and logistic regression | |
| CN105355038A (en) | Method for predicting short-term traffic flow through employing PMA modeling | |
| Peng et al. | A frequency-aware spatio-temporal network for traffic flow prediction | |
| CN111127879B (en) | Urban traffic flow prediction method based on generative countermeasure network | |
| CN115936242B (en) | Method and device for acquiring traceability relation data of air quality and traffic condition | |
| Slimani et al. | Road traffic forecasting using a real data set in Morocco | |
| Huang et al. | Towards automated model calibration and validation in rail transit simulation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191008 |